LexiSynth is an AI language assistant for live interactions. Multi-lingual transcription speech-to-text, translation and synthesis in real-time for streaming and content creation, locally on your computer or using cloud AI vendors. Works on Windows and Mac! (Currently in Beta - give us feedback! We are improving every day because of your help)
Download: https://github.com/occ-ai/lexisynth/releases
Features:
For usage support and requests: https://discord.gg/3EUBUjpCD3
Download: https://github.com/occ-ai/lexisynth/releases
- Transcription (speech to text) real-time local, with option for model size
- Translation real-time, local LLM and cloud (OpenAI, DeepL) options
- Multi-lingual: Both transcription and translation support >50 languages
- Output to .txt files for OBS integration
- Speech synthesis: Cloud (OpenAI, 11Labs) and local models
- Text output modality: .srt files, OBS websocket, direct HTTP stream to YouTube/Twitch live, HTTP server
- Audio output modality: application sound, audio files, NDI audio output
- Input modality: NDI audio input sources
For usage support and requests: https://discord.gg/3EUBUjpCD3